GB/T 6378.5-2025Sampling procedures for inspection by variables - Part 5: Sequential sampling plans indexed by acceptance quality limit (AQL) for inspection by variables (known standard deviation) (English PDF)

Foreword

This document complies with the provisions of GB/T 1.1-2020 "Standardization Work Guidelines Part 1: Structure and Drafting Rules of Standardization Documents".

Drafting.

This document is Part 5 of GB/T 6378, "Metrological Sampling Inspection Procedures". GB/T 6378 has the following parts published.

---Part 1: Lot-by-lot inspection of a single quality characteristic and a single AQL retrieved by Acceptable Quality Limit (AQL) plan;

---Part 3: Secondary Sampling Plans for Lot-by-Lot Inspection Retrieved by Acceptance Quality Limit (AQL);

---Part 4: Procedures for assessing the claimed quality level of the mean;

---Part 5: Sequential Sampling Inspection Plans Retrieved by Acceptance Quality Limit (AQL) (Standard Deviation Known);

---Part 7: Procedures for assessing the claimed quality level of nonconforming rates.

This document is modified to adopt ISO 3951-5:2006 "Metrological sampling inspection procedures – Part 5: Inspection of samples retrieved by acceptance quality limit (AQL)".

Sequential sampling inspection plan (standard deviation known).

The technical differences between this document and ISO 3951-5:2006, and the reasons therefor, are as follows.

---ISO 2859-1 has been replaced by GB/T 2828.1 (normative reference), and ISO 3534-1 has been replaced by GB/T 3358.1 (normative reference).

GB/T 3358.2 replaced ISO 3534-2, and GB/T 6378.1 replaced ISO 3951-1, to adapt to my country's technical requirements.

Add components to increase operability;

---The AQL corresponding to the code H in paragraph 4 of section 11.1 is at least 1.0 (%), therefore the AQL specification at test level S-3 is...

The statement "Setting it to 0.65% or lower is invalid" is revised to "The minimum AQL corresponding to the code H is 0.25%, therefore the test will be performed." the statement that "an AQL of 0.15% or lower at level S-3 is invalid" is incorrect.

---Modify "the intercept of the rejection line is -hR" and "the intercept of the receiving line is hA" in section 11.4.6.1 to "the intercept is -hRsigma" and "the intercept is hA" respectively.

The original text contains an error. "hAsigma".

---In section 11.4.9.2, "RL=gsigmancum-hR,Lsigma" should be changed to "RL=gLsigmancum-hR,Lsigma", as the original text contained an error.

---In section 11.4.9.2, "At,U=gLsigmant" should be changed to "At,L=gLsigmant", as the original text contained an error.

---In section 11.4.9.3.1, "When the cumulative sample size reaches the truncated sample size nt, if Y<=At,U, then the batch is not accepted, and the inspection is terminated."

---In section 11.4.9.3.2, "When the cumulative sample size reaches the truncated sample size nt, if Y >= At, L, then the batch is not accepted, and the inspection is terminated." the original text was incorrect; it should be corrected to "Y Introduction in modern production processes, a high-quality level of 10-6 defective products is often desired. In such cases, using conventional...

Sampling plans for acceptance testing by attributes (such as those provided in GB/T 2828.1) often require very large sample sizes. If measurement acceptance testing is used...

Sampling schemes (such as those provided in GB/T 6378.1) can significantly reduce sample size. However, this is especially true for received products.

Even with extremely high sample quality, the required sample size is still too large. Therefore, it is necessary to use standardized statistical procedures with the smallest possible sample size.

Sequential sampling is the only statistical sampling method that satisfies this need because it is the only one that can satisfy all possible sampling methods with similar statistical properties.

Among sampling schemes, sequential sampling schemes have the smallest average sample size. Therefore, it is essential to provide verification methods commonly used in GB/T 6378.1.

A sequential sampling test scheme that is statistically equivalent to a sampling scheme but requires a significantly smaller average sample size.

The main advantage of sequential sampling is the reduction in average sample size. Average sample size is a factor that must be considered within a given batch or process quality level.

The weighted average of all possible sample sizes in the sampling scheme is used. Under the premise of equivalent operational characteristics, sequential sampling is superior to single-stage sampling.

The average sample size of this sampling plan is smaller. The sequential sampling plan in this document introduces a truncation rule, limiting the upper limit of the actual number of units tested.

The value is 1.5n0, where n0 is the sample size of the corresponding single sampling scheme.

Other factors to consider include the following.

a) Complexity

Compared to the simple rules of a single sampling scheme, the rules of a sequential sampling scheme are slightly more complex.

b) Variability of the test quantity for a specific batch, since the actual number of units to be inspected is unknown in advance, the organization and implementation of the sequential sampling plan will present some challenges.

It's difficult. For example, the arrangement of inspection procedures.

c) Cost of sampling products If the cost of sampling products at different times is high, then the benefit of a reduced average sample size from a sequential sampling scheme may be negated.

The increase in sampling costs offset this.

d) Duration of the test If the testing time for a single product is long, and multiple products can be tested simultaneously, then the testing time using a sequential sampling scheme is shorter than that using a single sampling scheme.

Using the corresponding single sampling method takes a long time.

e) Intra-batch quality variation If a lot consists of two or more sub-lots from different sources, and there may be substantial differences in quality between the sub-lots, then sequential sampling is appropriate.

The random sampling method is more difficult to draw than the corresponding single sampling method.

Weighing the advantages of a small average sample size in sequential sampling schemes against the aforementioned disadvantages, we can conclude that sequential sampling schemes are only suitable for single...

The testing costs for sample products are relatively high.

The choice between single sampling and sequential sampling plans should be determined before the start of batch inspection. During a batch inspection, it is not permitted to...

The shift from one sampling plan type to another occurs because if the actual test results influence the choice of acceptance criteria, the operation of the sampling plan...

Its properties may change drastically.

Although sequential sampling is more economical on average than its counterpart single sampling, for the inspection of a specific batch, it may be less economical.

The cumulative difference (a statistic used to determine the acceptability of a batch) may remain between the accept and reject values for a long time, until the test quantity is very large.

Only in certain situations can a decision be made to accept or reject the application. When using the graphical method, the above situation corresponds to the step curve randomly oscillating within an indeterminate region.

To avoid the above situation, a truncated sample size nt is set for the cumulative sample size before sampling begins.

If the acceptability of a batch is not yet determined when the truncated sample size nt is reached, the test is terminated, and the truncated acceptance value and truncated rejection value are used to determine the batch acceptance.

Whether to accept or not.

For commonly used sequential sampling schemes, although truncation can alter the operational characteristics of the scheme, this document defines sequential sampling...

The operational characteristics of the sampling scheme take truncation into account; therefore, the truncation criterion is an integral part of the sampling scheme provided in this document.

GB/T 6378, the "Procedures for Metrological Sampling Inspection," is a series of fundamental and general standards concerning metrological sampling inspection. It is proposed that these standards be implemented by seven departments.

Composition.

---Part 1: Lot-by-lot inspection of a single quality characteristic and a single AQL retrieved by Acceptable Quality Limit (AQL) the plan specifies the acceptance sampling system for a single-sample inspection plan in metrology, indexed by the Acceptable Quality Limit (AQL). (Headquarters) the purpose of the proposed method is to ensure a high probability of acceptance for batches whose actual quality level is better than the acceptance quality limit, while also ensuring...

Batches whose actual quality level is inferior to the acceptance quality limit will be accepted with a low probability.

---Part 2: General Rules for Single Sampling Plans for Lot-by-Lot Inspection of Independent Quality Characteristics Retrieved by Acceptance Quality Limit (AQL) this paper defines a broader technical approach for lot-by-lot measurement sampling inspection of independent quality characteristics.

---Part 3: Lot-by-lot sampling plans for inspection retrieved by Acceptable Quality Limit (AQL). This specifies the metrological secondary sampling inspection procedures.

The acceptance sampling system for the plan is a supplement to GB/T 6378.1.

---Part 4: Procedures for assessing the claimed quality level of the mean. This specifies the procedures for assessing the average of a certain quality characteristic of the population being examined.

This is a single sampling inspection procedure for measurement quality indicators, applicable to various forms of quality verification, and is part of the GB/T 6378 series.

The procedure used to assess claimed quality levels.

---Part 5: Sequential Sampling Inspection Plans Retrieved by Acceptance Quality Limit (AQL) (Standard Deviation Known). This is...

Supplement to GB/T 6378.1.

---Part 6: Sampling Plan for Isolated Lots Retrieved by Limiting Quality (LQ). This sampling system is used for isolated lots (isolated sequences).

The batch (isolated batch or single batch) inspection is a supplement to GB/T 6378.1 and is compatible with GB/T 6378.1.

---Part 7: Procedures for assessing the claimed quality level of nonconforming rates. This is a supplement to GB/T 6378.4, and...

GB/T 6378.4 together constitute the GB/T 6378 series of procedures for assessing claimed quality levels.

Part 5 of the Metrological Sampling Inspection Procedure. By Acceptance Quality Sequential sampling of measurement based on the AQL (Advanced Quality Limit) retrieval Test plan (standard deviation known)

1 Scope

This document specifies the acceptance sampling system for lot-by-lot sequential sampling inspection plans. These acceptance sampling plans are based on a series of...

Priority is given to using Acceptable Quality Limit (AQL) values for retrieval, ranging from 0.01 to 10.AQL is defined based on the percentage of nonconforming items. Using this...

When creating the document, it is necessary to ensure that the production process is stable (under statistical control).

The purpose of this acceptance sampling plan is to exert economic and psychological pressure on the manufacturer by rejecting substandard batches, thereby prompting them to...

The process average quality level is kept below the specified acceptance quality limit, while providing an upper limit on the probability that the user will accept a substandard batch.

These acceptance sampling plans apply to consecutive lot series, i.e., lot series of sufficient length to allow the use of the transfer rules in Chapter 6.

Then, these transfer rules are.

---Once a deterioration in quality is detected, protection for the user will be strengthened by transferring the process to more stringent inspections or suspending sampling inspections;

---If the quality remains consistently good, to reduce inspection costs, the responsible department has decided, according to the transfer rules, to transfer to a less stringent inspection facility.

Provide incentives for producers.

The sampling plan in this document is designed based on the following conditions being met.

a) The inspection procedures in this document apply to consecutive batches of separate individual products, i.e., all products are manufactured by the same producer and produced in the same manner.

This document is provided by the process. If there are different producers or production processes, this document shall be applied to each producer or production process separately.

b) Consider only a single quality characteristic x of the product, which must be measured using a continuous scale.

c) The measurement system uncertainty is negligible relative to the standard deviation of the production process.

d) The production process is stable (statistically controlled), and the quality characteristic value x follows a normal distribution (or approximately follows a normal distribution) or the distribution can be transformed into a normal distribution.

e) The standard deviation of the quality characteristic x is known.

Note. The sampling inspection procedures in this document are not applicable to batches in which nonconforming products have been pre-screened.

f) If the contract or standard specifies an upper specification limit U, or a lower specification limit L, or both, for the product quality characteristics; and A product is considered acceptable only if the measured quality characteristic x satisfies one of the following inequalities.

1) Quality characteristic value x<=U (i.e., not higher than the upper specification limit).

2) Quality characteristic value x >= L (i.e., not lower than the lower specification limit).

3) Quality characteristic value L<=x<=U (that is, the quality characteristic value is between the lower specification limit and the upper specification limit).

If two-sided specification limits are applied, this document assumes that both specification limits are equally important for the product's conformity, or assumes that both specification limits are equally important for the product's conformity.

Consider them separately. In the first scenario, it is advisable to specify only one total nonconforming quantity (AQL) for both sides, i.e., the "joint AQL requirement".

In the second case, it is advisable to specify the percentage of nonconforming items (AQL) for each side of the two-sided specification limit, i.e., "separate AQL requirements".

In this document, the acceptability of a batch is essentially determined by the percentage of process nonconforming items estimated from a randomly drawn sample of the batch.

For isolated batches or consecutive batches whose sequence length does not meet the requirements of this document, users are advised to consult GB/T 2828.2 for appropriate guidance.

Sampling plan.

2 Normative references

The contents of the following documents, through normative references within the text, constitute essential provisions of this document. Dated citations are not included.